Micromorphological observations from 3 different published works have been studied to aid understanding of aggregation and of colloids, both unique to soils. Saprolites in Hong Kong included ‘veins’ of different thicknesses and colours. Optical mineralogy identified them as infill from the neogenesis of clays in rock fractures.
Turkish Journal of Earth Sciences http://journals.tubitak.gov.tr/earth/ Research Article Turkish J Earth Sci (2013) 22: 376-390 © TÜBİTAK doi:10.3906/yer-1201-11 The key role of micromorphology in studies of the genesis of clay minerals and their associations in soils and its relevance to advances in the philosophy of soil science Gordon Jock CHURCHMAN* School of Agriculture, Food and Wine, Waite Campus, The University of Adelaide, Private Mail Bag No.1, Glen Osmond, South Australia 5064, Australia Received: 31.01.2012 Accepted: 23.04.2012 Published Online: 06.05.2013 Printed: 06.06.2013 Abstract: Micromorphological observations from different published works have been studied to aid understanding of aggregation and of colloids, both unique to soils Saprolites in Hong Kong included ‘veins’ of different thicknesses and colours Optical mineralogy identified them as infill from the neogenesis of clays in rock fractures The common thicker infills resulted from weathering Dark infill contained comminuted primary minerals whereas thin pale infill originated hydrothermally Scanning electron microscopy (SEM) showed that the size, shape, and mineralogy of the kaolin minerals formed in infill depended on the types of cracks in the saprolites and on drying Energy-dispersive X-ray spectroscopy analyses showed Fe and/or Mn in dark-coloured infill from comminution of primary minerals upon brecciation, or else beside pale infill in tuff, showing seasonal drying in tuff but not in granite Pale infill gave predominantly large tubular halloysite in granite but large platy kaolinite in tuff, except that hydrothermal kaolin gave small particles In dark infill, kaolin particles were also small and were kaolinite and halloysite mixtures The effect of impurity Fe and Mn in constraining kaolin mineral crystallinity in infills simulates some of the effects of impure soil environments Long-term cultivation of soils in Australia led to environmental scanning electron microscope images of large microaggregates indicating their breakdown and loss Transmission electron micrographs of ultrathin sections showed that microaggregates of clay size, comprising clay minerals and oxides covering other materials, including organic matter, were predominant in virgin soil but were broken down to fine clay particles that blocked pores in cultivated soils SEM showed a web of biological origin in long-term irrigated sandy New Zealand soil that surrounded macroaggregates but only became closely attached on drying The nature of the macroaggregates was affected strongly by their history of drying, even during preparation for analyses Micromorphology is especially useful for indicating the nature of aggregates in situ in soils Key Words: Aggregates, microaggregates, macroaggregates, aggregate stability, electron microscopy, colloids, neogenesis Introduction Micromorphology is an established sub-discipline of soil science Its foundation probably lies in the use of hand lenses for magnifying the features of soils in the field, hence expanding the view available to the naked eye Thin sections have been studied under optical microscopes for the understanding of soil genesis since the beginning of the 20th century (Stoops 2010), but Stoops (2010) considers that the study of micromorphology had its real start with the publication of W.L Kubiëna’s book Micropedology in 1938 Any study of the fine-level structures or morphology visible through microscopy, including those of non-soil materials, can be strictly characterised as micromorphology Since the early 20th century, the scope of microscopy has advanced dramatically, mainly through the use of electron optical methods Because the unique contribution of micromorphology to studies of soils * Correspondence: jock.churchman@adelaide.edu.au 376 and other natural objects comes from its ability to view these objects in situ, thereby minimising artefacts from their preparation, scanning electron microscopy (SEM) is the electron optical approach that has been used most commonly in these studies SEM continues to be widely used in soil studies, for minerals (e.g., Churchman et al 2010b), organic materials, and also their associations (e.g., Miltner et al 2011) Its use with an environmental cell (as environmental scanning electron microscopy, or ESEM) means that any effects of strong drying beyond that experienced by soils in nature can be avoided during the preparation of soils for viewing This is especially advantageous for studying biological entities in soils, as well as, potentially, for some soil aggregates (Foster 1994; Churchman et al 2010a) Transmission electron microscopy (TEM) has also been used, especially by R.C Foster in Australia and C Chenu in France, to study soils using preparative techniques that leave material, in CHURCHMAN / Turkish J Earth Sci ultrathin sections, largely physically intact for viewing (e.g., Foster & Martin 1981; Chenu 1989; Chenu & Plante 2006; Churchman et al 2010a) In this study, published work, largely by me and co-workers, is used to illustrate some of the uses of micromorphology at different scales to solve problems relating to the genesis of some soils and soil minerals and also to the nature of associations between soil minerals and other components in some other soils The main use of these examples herein is to point to an important role that micromorphology may be able to play in advancing our philosophical understanding of soils Probably the major advantage of the various micromorphological tools for the study of soils is that they can provide views of the soils in situ, as already discussed Many methods of studying soils and their components require chemical and physical pretreatments that produce artefacts comprising materials that may have lost some of the defining characteristics that constitute soils as a unique object of study Therefore, micromorphological studies potentially have a key role to play in understanding the unique and important characteristics of the materials we call soils This study is mainly concerned with the contributions that micromorphology can make to discovering the characteristics of soils that make them unique among materials for scientific study Micromorphological studies by their very nature have also made, and continue to make, contributions to discerning important characteristics of soils It may be argued that the most useful explanation in soils reside at the level of plant roots, biota (including microbes), and water and nutrients Explanations at the atomic level are not of much use in soils (e.g., Churchman 2010a) Furthermore, roots, biota, and water are concerned with aggregated soil, not with crushed, disaggregated, or even dried soil The strength of micromorphological studies is that they observe aggregated, and largely undisturbed, soil Hence, this study seeks to ascertain the role that micromorphology, using optical, electron-optical, and also newer techniques such as those using X-ray microscopy (e.g., Wan et al 2007) and computer-assisted tomography (Tracy et al 2010), may be able to play in better defining soils as a philosophical entity The philosophical framework for the study was established by Churchman (2010a) According to Churchman’s (2010a) analysis, soils have aspects that mark them as unique objects of study These are: (i) the formation and properties of horizons, (ii) the occurrence and properties of aggregates, and (iii) the occurrence and behaviour of unique colloids Respectively, these may be defined as the unique macro-, micro-, and nano-characteristics of soils It is already evident from the literature that each of these has been the subjects of study by micromorphological techniques In the pedological context, micromorphological studies, generally at the macro-level using optical microscopy, have been carried out on different horizons of soils The micromorphology of distinctive horizons including gypseous, spodic, mollic, takyric, and yermic, as well as the commonly named A, B, and C horizons, has been the topic of many studies (see, for example, many of the chapters in Stoops et al (2010)) Characterisation of their micromorphological features has enhanced the understanding of their genesis and that of their constituent soils In this study however, emphasis is given to studies of aggregates and colloids at the micro- and nano-levels, respectively Outline of the studies The micromorphological results from studies are presented here The studies are: Saprolite weathering, Hong Kong (Churchman et al 2010b) Among micromorphological techniques, this study employed mainly optical microscopy of this section and SEM of whole (rock) samples Long-term effects of agriculture on an Alfisol soil, South Australia (Churchman et al 2010a) This study employed the micromorphological techniques of ESEM of intact aggregates separated from soils and TEM of ultrathin sections of resin-embedded sections of whole soils Effects of irrigation on an Inceptisol, New Zealand (Churchman & Tate 1986) This study employed only SEM for micromorphology Most of the details of the setting of the samples and preparative techniques can be found in the references cited, but some are summarised and illustrated herein under ‘Materials’ Materials 3.1 Saprolite weathering Since the project including this study was carried out with the major objective of explaining the role played by kaolin-rich vein-like zones within saprolites on slopes in Hong Kong in causing or enhancing landslides, the study mainly focused on samples comprising these ‘veins’ The saprolites have formed within either granite or volcanic tuff as a result of weathering under a very high rainfall It had been established that they could include either or both halloysite or kaolinite and therefore their analysis was able to add to our understanding of the conditions under which halloysite or kaolinite were formed authigenically from the products of weathering of granite or volcanic tuff Figure shows a kaolin-rich ‘vein’ within volcanic tuff on a slope in Hong Kong For the study, block samples of approximately 100 × 100 × 50 mm in size were collected from the saprolites at 377 CHURCHMAN / Turkish J Earth Sci Cul tiva te > Newly cultivated 100 yea rs Churchyard (virgin) Trial plots, 250 m (Cultivated > 100 years + NT, CT, 18 years) Figure A photograph of saprolite from the weathering of volcanic tuff on a slope and an incorporated white vein-like feature that is shown at the true angle to the slope The width of the ‘vein’ ranges up to approximately 10-20 mm 20 sites, 10 of them from granite and 10 from volcanic tuff, and were transported to the laboratory without drying While some sub-samples were removed from ‘vein’ and surrounding material on the blocks for SEM and other studies, the largest part of the block was impregnated with a resin following air-drying and thin sections were cut for optical microscopy SEM was conducted with an energy dispersive X-ray (EDX) detector Samples were coated with gold for SEM imaging and with carbon for EDX analyses 3.2 Long-term effects of agriculture In this study, samples of the same soil type, which had been subjected to common, and sometimes also experimentally controlled, agricultural practices over periods of time of up to approximately 120 years, were compared for the effects of these practices on the nature of the soils, and particularly on the associations between their constituents The study was enabled by the availability of a virgin site adjacent to a recently cultivated and farmed site, also quite close to rotation and tillage trial sites located on land that had been formed for ca 100 years The site of the virgin soil was located within a plot of land that had been occupied by a church building from the beginning of the settlement of this region in 1869 until 1949 and which had remained fenced off and never cultivated since The terrain is quite flat over the area comprising all sites The area including the virgin site and the recently cultivated site, and also the location of the trial sites, are shown in an aerial photograph in Figure Generally, samples for micromorphological analyses were taken from cores removed from the soils at intervals ranging from 0.01 m at the tops of the profiles to >0.1 m at greater depths 3.3 Effects of irrigation The availability of sites, about km apart, on the same sandy soil type, where soil had been irrigated with effluent 378 Figure An aerial photograph (from Google Earth, taken December 2006) showing the sampling spots (numbered) in the churchyard site of the virgin soil and the adjacent newly cultivated soil as well as the surrounding soil that has been cultivated for >100 years The site of the plots in which tillage (including no-tillage, NT, and conventional tillage, CT) and crop rotation trials were carried out for 18 years following cultivation for >100 years overall is indicated, although outside of the view shown Reproduced from Churchman et al (2010a) with permission from Elsevier from an abattoir and kept moist for 25 years at one site while it had been irrigated with water to maintain a 20% moisture content at an irrigation research station at the other site for 30 years, enabled this study The soils were maintained under permanent grass-clover pasture, which was grazed by sheep or cattle There were control sites at each site and these both dried out each summer The main object of the study had been to determine the effect of the disposal of the abattoir effluent upon aggregation in the soil, and the inclusion of the soil which had been irrigated with water alone for a similar period of time was aimed to enable the separation of the effects of water alone in the abattoir effluent from that of the water inevitably added along with this effluent SEM was carried out on 3.4-2.0 mm aggregates separated from the soils by wet sieving The aggregates were examined by SEM both before airdrying and after freeze-drying, and also after air-drying Results and interpretations 4.1 Saprolite weathering While optical microscopy was carried out on both matrix and ‘vein’ material, the most useful information was obtained from the latter Nonetheless, it was observed that kaolin alteration was ubiquitous and extensive throughout the host rocks studied In saprolites from both granite and volcanic tuff, feldspars showed the greatest degree of alteration Alteration of biotite and sometimes also of muscovite was observed in the matrix of the saprolite, although some muscovite remained unaltered Quartz appeared to be unaltered throughout CHURCHMAN / Turkish J Earth Sci The ‘veins’ varied in colour from white through pink, shades of yellow, and brown, and, in some cases, were black Their colours have been identified more objectively using the Munsell scheme (see Churchman et al 2010b) Even so, they could be separated into pale or dark The textures also varied, ranging from clayey to sandy silt Pale veins were either clay or silty clay in texture, while dark veins covered a wider range, including the coarser grades of sandy, silty clay, and sandy silt Veins also varied in thickness or width between samples, but were generally >10 mm at their thickest in any one sample, although some were as thick as 55 mm They also varied in thickness within samples, as seen in Figures 3-7 In samples, both in saprolite from tuff at the FNS (Fei Ngo Shan) locality, the white veins were notably narrow; they were always narrower than mm A further point of distinction between the veins in these samples and those from all other samples was that those in FNS occurred as broad networks of intersecting veinlets, characterised as ‘box- work’, in stark contrast to each of the veins in all other 18 samples, which were in a parallel or sub-parallel alignment with other veins where they occurred in the same sample This distinction pointed to a genetic difference that was explored (see below) between the origin of the veins in the FNS samples and those at all other sites Overall, the nature of the kaolin clay minerals – and other minerals – occurring in the veins appeared to have a direct association with the thickness and colour of the veins, although thick white veins differed also according to their lithologies, whether granitic or tuffaceous Samples were therefore separated into types according to the thickness and colour of their included veins These types and their optical analyses, as well as their clay mineralogies, from SEM were as follows: Thick white veins (a) in granite: These are represented in Figure by sample TKL2 from Tiu Keng Leng (b) in tuff: These are represented in Figure by sample SSR DS1 from Sai Sha Road 2200 4079 s6 ug10.5 flow lhs Si 2000 Al Intensity 1800 1600 1400 Au-M 1200 1000 800 600 400 200 K K Mn Fe Au-L Fe 10 Energy (KeV) Figure Top left: Block sample TKL2 (approx 100 mm2), showing thick white vein towards top of sample Top right: Microfabric of white vein within sample TKL2; the scale bar is mm long Lower left: SEM of white vein in TKL2 at low magnification Lower right: EDX analysis of the vein in TKL2 Partly reproduced from Churchman et al (2010b) with permission from the Clay Minerals Society 379 CHURCHMAN / Turkish J Earth Sci 1400 4157 s01 ug342 yellow average Si 1200 Al Intensity 1000 800 600 Au-M 400 200 K K Ti Ti MnFe Fe Au-L 10 Energy (KeV) Figure Top left: Block sample SSR DS1 (approx 100 mm2), showing thick white vein towards top of sample, as well as thinner black veins and also black spots Top right: Microfabric of white vein (top) and also thinner black vein within sample SSR DS1; the scale bar is mm long Lower left: SEM of white vein in SSR DS1 at low magnification Lower right: EDX analysis of the vein in SSR DS1 Partly reproduced from Churchman et al (2010b) with permission from the Clay Minerals Society Thin white veins These are represented in Figure by sample FNS N from Fei Ngo Shan Thick brown veins These are represented in Figure by sample TKL3 from Tiu Keng Leng Thick black veins These are represented in Figure by sample STC S1A from Sha Tin College The major features of the images and analyses in Figures 3-7 and the others of similar types that they represent (Churchman et al 2010b) that require explanation include: The reason why veins are either ‘white’ (or other light colours such as pink) or dark, including shades of brown, yellow, red, or black The reason for the different sizes and shapes of claysized particles in SEM 380 The reason why the veins in samples from one location (FNS) are much thinner than the veins in samples from other locations and that they have a unique random or box-work configuration among the other samples in the study The explanations are detailed by Churchman et al (2010b) but, in summary, they are explained by the origin of the veins Fresh rock, whether granite or tuff, has undergone alteration on the slopes This has occurred either by weathering, or by hydrothermal alteration Alteration has led to the replacement of the most easily altered primary minerals by secondary minerals X-ray diffraction analyses, as well as previous studies on samples from the slopes of Hong Kong (Kirk et al 1997; Campbell CHURCHMAN / Turkish J Earth Sci 4304 s02 ug65.1 top slicken Si 1200 1000 Intensity 800 600 Au 400 Al 200 K Mg Na Au Fe Ca Ti Mn Fe 10 Energy (KeV) Figure Top left: Block sample FNS N (approx 100 mm2), showing thin white veins throughout sample Top right: Microfabric of white vein within sample FNS N; the scale bar is mm long Lower left: SEM of white vein in FNS N at high magnification Lower right: EDX analysis of the vein in FNS N Partly reproduced from Churchman et al (2010b) with permission from the Clay Minerals Society et al 1998) indicated that either halloysite or kaolinite constituted the bulk of the secondary minerals formed The rocks have become weakened as a result of alteration of their constituent minerals Especially because of the load imposed by materials upslope, the weakening of the rocks has led to their fracture This has occurred either along intergranular contacts within the rocks or else by shearing of crystals Fracturing that occurred along intergranular contacts would lead to clean, uncontaminated fracture surfaces between rock fragments while that occurring with shearing of crystals would lead to a brecciation of these crystals The brecciation would lead to the comminution of primary minerals into finer fragments and hence to their easier dissolution, to give especially oxides and hydroxides of iron and manganese The veins are explained generally by the neogenesis of kaolin minerals from solutions that have leached from the rocks during their alteration They are more correctly described as ‘infill’ On the basis of this genetic mechanism, the explanations of the particular features of Figures 3-7 identified here are proposed as follows: Colour of infill Infill is white when rock fracture has occurred largely along intergranular contacts, leaving clean surfaces for neogenesis to occur in the newly formed void, devoid of coloured contaminants This is so in the representative samples described in Figures 3-5 The EDX analyses in Figure show almost no peak for the colouring elements Fe and Mn Apart from that for the covering Au, the analyses are dominated by those for Al and Si, with Si > Al, consistent with the composition of the kaolin minerals That in Figure is similar but shows very small peaks for Fe and Mn, as well as minor peaks for K and Ti The analyses in Figure are essentially the same, although there are significantly stronger peaks for Fe, especially, along with small peaks for K and Ti, in this case (sample FNS N) These may arise from the bulk of 381 CHURCHMAN / Turkish J Earth Sci 4083 s10 ug14.1 smooth mat Si 6000 Intensity Al 5000 4000 Au-M 3000 2000 Fe 1000 KK Ti Mn Au-L Fe 10 Energy (KeV) Figure Top left: Block sample TKL3 (approx 100 mm2), showing brown veins throughout the sample Top right: Microfabric of brown vein within TKL3; the scale bar is 0.25 mm long Lower left: SEM of brown vein in TKL3 at intermediate magnification Lower right: EDX analysis of the vein in TKL3 Partly reproduced from Churchman et al (2010b) with permission from the Clay Minerals Society this sample bordering especially narrow infill (Figure 5), as already noted The resolution of the beam for analysis may be insufficiently small to include just infill materials so that primary minerals such as K-feldspar and titanium oxide contribute to the analyses The dominant infill in the samples shown in Figures 3-5 is largely monochrome, although there are textural differences, especially between that in FNS N (Figure 5) and those in TKL2 (Figure 6) and SSR DS1 (Figure 7), as will be explained further below The black infill alongside the dominant white infill in SSR DS1 (Figure 6) has another origin (see below) By contrast, coloured infill may include considerable Fe, as in TKL3, and this contributes to the various shades of red, yellow, and brown in the infill in this sample (Figure 6) 382 and/or Mn, which is largely responsible for the dominantly black infill in STC S1A (Figure 7) K and Ca are also present in notably high proportions, indicating the incorporation of substantial primary minerals in the infill in this sample (STC S1A) The optical micrograph for TKL3 (Figure 6) and STC S1A (Figure 7) shows that infill in these samples is very heterogeneous in terms of colour, at least That for STC S1A also shows great heterogeneity, and also a high concentration of small comminuted particles that have resulted from the brecciation of primary minerals upon rock fracture occurring within mineral grains There is a huge difference between the sizes of the dominant particles in the different infills Those shown in the SEMs in Figures and within thick white infills CHURCHMAN / Turkish J Earth Sci 1200 4207 s13 u41.1 fibres Mn 1000 800 Intensity Au Si 600 Al 400 Mn Fe K 200 Ca Au Fe 10 Energy (KeV) Figure Top left: Block sample STC S1A (approx 100 mm2), showing black veins throughout the sample Top right: Microfabric of black veins within STC S1A; the scale bar is mm long Lower left: SEM of black vein in STC S1A at intermediate magnification Lower right: EDX analysis of the vein in STC S1A Partly reproduced from Churchman et al (2010b) with permission from the Clay Minerals Society are large, although they differ from each other in their dominant shape They comprise very long tubular particles in TKL2 infill (Figure 3) and quite large platy particles, assembled together in the shape of rosettes, in SSR DS1 infill (Figure 4) In these, and in all other samples, X-ray diffraction (XRD) analyses have identified tubular particles as halloysite and platy particles as kaolinite In TKL2 and SSR DS1, differential thermal analyses (DTAs) showed that kaolin minerals comprised at least 80% of the infill (Table in Churchman et al 2010b) XRD showed that 100% of the kaolin minerals in TKL2 infill are halloysite, while 80% of them in SSR DS1 infill are kaolinite (Tables and in Churchman et al 2010b) By contrast, the particles of kaolin minerals in the infills in FNS N, which is white, and in both TKL3 and STC S1, which are highly coloured, are much smaller than those in TKL2 and SSR DS1 They appear to be highly tubular in TKL3 (Figure 6), platy in FNS N (Figure 5), and a mixture of shapes in STC S1 (Figure 7) XRD analysis confirmed abundant halloysite in TKL3, although kaolinite was present in nearly as high a concentration, and it indicated a significantly higher concentration for halloysite than for kaolinite in FNS N This confirms that electron microscopy is probably too selective and/or misleading when one shape (tubular in this case) is visually dominant for good quantitative analyses For STC S1A, DTA shows that the proportion of infill that comprised kaolin minerals was very low XRD showed a crystalline manganese oxide, todorokite, to be present and other SEM images showed this to comprise quite large platy particles 383 CHURCHMAN / Turkish J Earth Sci (a) (b) (c) (a) (b) (c) Figure Environmental scanning electron micrographs (ESEMs) (left) and transmission electron micrographs (TEMs) of ultrathin sections (right) of samples from within the upper 0.05 m of (a) virgin soil, (b) newly cultivated soil adjacent to virgin soil site, and (c) soil under long-term conventional cultivation Scale bars represent 50 µm in ESEMs and µm in TEMs “Q” indicates grains of quartz and “qz” indicates quartz shards “M” indicates microaggregates; “cl”, clay within microaggregates; “fc”, fine dispersed clay outside of microaggregates; “om”, organic matter Reproduced from Churchman et al (2010a) by permission from Elsevier Therefore, the SEM in Figure shows tubular halloysite, platy kaolinite, and also platy todorokite The evidently substantial occurrence of the latter is consistent with the high proportion of Mn shown in the EDX analysis of this sample (Figure 7) The explanation for the comparatively larger sizes of particles in TKL2 and SSR DS1 (Figures and 4) than in others lies in the relatively clean environment (open cracks) in which kaolin minerals formed by neogenesis in these samples The reason why kaolin minerals formed in the coloured infills are small comes from the constraints 384 that the other ions in solution (those of Fe and/or Mn, mainly) imposed upon crystal growth in the contaminated environments resulting from the brecciated fractures The explanation why halloysite is formed rather than kaolinite, or vice versa, in the various samples of infill was suggested by the appearance of manganese oxide, as black veins or black spots, and/or iron oxides in or alongside the infill many of the samples, especially SSR DS1 (Figure 4), TKL3 (Figure 6), and, of course, STC S1A (Figure 7) Only those samples containing infill including or bordering on black spots or veins of manganese oxide and/or red, yellow, CHURCHMAN / Turkish J Earth Sci a b c d Figure Scanning electron micrographs (SEMs) of the surfaces of macroaggregates of 2-3.4 mm in size from a soil: (a) (top left) Irrigated with water to 30% moisture content for 30 years; aggregate studied freeze-dried (b) (top right) From control site adjacent to water-irrigated soil; aggregate studied freeze-dried (c) (lower left) Irrigated with effluent from an abattoir and kept moist for 25 years; aggregate studied freeze-dried (d) (lower right) Irrigated with water to 30% moisture content for 30 years; aggregate studied air-dried Reproduced from Churchman and Tate (1986) by permission from CSIRO Publishing or brown colouring from iron oxides or oxyhydroxides contained kaolinite Otherwise, where these features did not appear, the kaolin minerals in infill were predominantly halloysite The white infill in TKL2 (Figure 3) contains only halloysite among the kaolin minerals Manganese and iron oxides or oxyhydroxides both require drying for their formation, so it is concluded that their occurrence indicates that the infills containing or bordering these oxides have undergone periods of drying Halloysite is formed in its hydrated state (Churchman & Carr 1975), so it can be concluded that, when drying occurs, kaolinite is favoured as the newly formed kaolin mineral, whereas halloysite only forms when the environment remains wet Drying is only intermittent, and probably seasonal, in the high rainfall zone of Hong Kong, so it appears that mixtures of halloysite and kaolinite, such as in all samples examined here except TKL3, result from different hydration regimes occurring cyclically in the corresponding sites The exceptional infill in FNS samples Both the optical evidence and that from SEM suggest that FNS samples, represented by FNS N (Figure 5) here, have a different origin from the other samples in this study The microfabric by optical microscopy in Figure appears to be unstressed, showing randomly disposed microvermiform shapes, unlike those in the other samples in Figures 3, 4, 6, and 7, which reflect processes of shearing and/or brecciation having taken place in their formation and development The clay particles also differ in their 385 CHURCHMAN / Turkish J Earth Sci a b c Figure 10 Transmission electron micrographs (TEMs) of ultrathin sections (right) of samples from within the upper 0.05 m of an Alfisol from South Australia, showing dark mineral matter surrounding (a) (top) plant cells, probably fine roots (pale), and also quartz shards; (b) extracellular polysaccharide (identified by staining); and (c) bacteria The scale bar in each case represents µm Partly reproduced from Churchman (2000) with permission from CRC Press alignment and association with one another from the other samples studied They comprise relatively small particles, which are randomly interlocked and reasonably tightly packed These characteristics mark them as typical products of hydrothermal processes, according to Keller (1976) Weathering, by contrast, tends to give looser arrangements of particles (Keller 1976) A hydrothermal origin is possible in Hong Kong, especially close to fault zones (Irfan 1996) The infill in the FNS samples may not have formed by complete fractures of rocks as in the other samples studied 4.2 Long-term effects of agriculture In this study, described in detail by Churchman et al (2010a), micromorphology was carried out using ESEM on air-dried aggregates and with TEM, which was applied to ultrathin sections of resin-treated samples of topsoil While many other measurements were made on the soil samples, 386 micromorphology proved crucial in understanding the effects on the soil studied of different extents of agricultural management The micromorphological data enabled explanations of the differences between soils with different histories that were recorded in particle and pore size distributions, surface spectral analyses, and various measures of aggregate stability to be made in terms of observable changes in the nature and extent of aggregation Examples of observations made with ESEM and TEM of soils with different land-use histories are given in Figure The ESEMs in each case (virgin, newly cultivated, and long-term cultivated soils) show more-or-less rounded microaggregates that are each just a few micrometres (generally